With the latest advances in information and communication technologies, greater amounts of sensitive user and corporate information are shared continuously across the network, making it susceptible to an attack that can compromise data confidentiality, integrity, and availability. Intrusion Detection Systems (IDS) are important security mechanisms that can perform the timely detection of malicious events through the inspection of network traffic or host-based logs. Many machine learning techniques have proven to be successful at conducting anomaly detection throughout the years, but only a few considered the sequential nature of data. This work proposes a sequential approach and evaluates the performance of a Random Forest (RF), a Multi-Layer Perceptron (MLP), and a Long-Short Term Memory (LSTM) on the CIDDS-001 dataset. The resulting performance measures of this particular approach are compared with the ones obtained from a more traditional one, which only considers individual flow information, in order to determine which methodology best suits the concerned scenario. The experimental outcomes suggest that anomaly detection can be better addressed from a sequential perspective. The LSTM is a highly reliable model for acquiring sequential patterns in network traffic data, achieving an accuracy of 99.94% and an f1-score of 91.66%.

Adversarial attacks pose a major threat to machine learning and to the systems that rely on it. In the cybersecurity domain, adversarial cyber-attack examples capable of evading detection are especially concerning. Nonetheless, an example generated for a domain with tabular data must be realistic within that domain. This work establishes the fundamental constraint levels required to achieve realism and introduces the adaptative perturbation pattern method (A2PM) to fulfill these constraints in a gray-box setting. A2PM relies on pattern sequences that are independently adapted to the characteristics of each class to create valid and coherent data perturbations. The proposed method was evaluated in a cybersecurity case study with two scenarios: Enterprise and Internet of Things (IoT) networks. Multilayer perceptron (MLP) and random forest (RF) classifiers were created with regular and adversarial training, using the CIC-IDS2017 and IoT-23 datasets. In each scenario, targeted and untargeted attacks were performed against the classifiers, and the generated examples were compared with the original network traffic flows to assess their realism. The obtained results demonstrate that A2PM provides a scalable generation of realistic adversarial examples, which can be advantageous for both adversarial training and attacks.

Cyber-physical systems are infrastructures that use digital information such as network communications and sensor readings to control entities in the physical world. Many cyber-physical systems in airports, hospitals and nuclear power plants are regarded as critical infrastructures since a disruption of its normal functionality can result in negative consequences for the society. In the last few years, some security solutions for cyber-physical systems based on artificial intelligence have been proposed. Nevertheless, knowledge domain is required to properly setup and train artificial intelligence algorithms. Our work proposes a novel anomaly detection framework based on error space reconstruction, where genetic algorithms are used to perform hyperparameter optimization of machine learning methods. The proposed method achieved an F1-score of 87.89% in the SWaT dataset.

Critical infrastructures are an attractive target for attackers, mainly due to the catastrophic impact of these attacks on society. In addition, the cyber–physical nature of these infrastructures makes them more vulnerable to cyber–physical threats and makes the detection, investigation, and remediation of security attacks more difficult. Therefore, improving cyber–physical correlations, forensics investigations, and Incident response tasks is of paramount importance. This work describes the SMS-I tool that allows the improvement of these security aspects in critical infrastructures. Data from heterogeneous systems, over different time frames, are received and correlated. Both physical and logical security are unified and additional security details are analysed to find attack evidence. Different Artificial Intelligence (AI) methodologies are used to process and analyse the multi-dimensional data exploring the temporal correlation between cyber and physical Alerts and going beyond traditional techniques to detect unusual Events, and then find evidence of attacks. SMS-I’s Intelligent Dashboard supports decision makers in a deep analysis of how the breaches and the assets were explored and compromised. It assists and facilitates the security analysts using graphical dashboards and Alert classification suggestions. Therefore, they can more easily identify anomalous situations that can be related to possible Incident occurrences. Users can also explore information, with different levels of detail, including logical information and technical specifications. SMS-I also integrates with a scalable and open Security Incident Response Platform (TheHive) that enables the sharing of information about security Incidents and helps different organizations better understand threats and proactively defend their systems and networks.